Beef Improvement Federation
Impact of Single Step on Selection Indexes
By Lindsay King
Implementing single-step methodology
was a big task, but now the real job is
showing the value in the headache of
it all. Matt Spangler, associate professor
of animal science at the University
of Nebraska-Lincoln, discussed the
far-reaching influence single step has
and will have on selection indexes.
“In a nut shell, they (selection
indexes) are a tool to enable
informed multiple-trait selection.
Every bull buyer does this in some
manner already,” Spangler said. “This
is a much more comprehensive
and informed way to do it.
“Generally, the indexes we have
now are static,” he continued.
“As we customize these tools, it
allows us to take advantage of the
phenotypic enterprise means.”
The goal is to improve commercial
profitability. The economic models
used before are proving correct
on average, but the single-step
methodology takes what the producers
value most into consideration
when developing the index. have to re-evaluate the economic
value we originally assigned to it.”
While relatively new to the beef
industry (rolled out about 10 years
ago), selection indexes have been used
in other industries since 1942, Spangler
said. “It is old hat to other industries
— the idea of selecting candidates
to be parents not just on selection
indexes, but exclusively from those.” Other factors affecting the
index include component trait
accuracy, economic parameters
and assumptions and assumed
population means. Some make the
mistake of blaming single step for
inaccuracy or unwanted results after
changing too many factors at once.
Many things have an effect on
selection indexes — the goal traits
that a producer is looking for in a
program is the main factor. However,
expected progeny differences (EPDs)
are always changing. As more data
comes in and new EPDs are developed
for a trait that is economically
relevant, these selection indexes
can be altered continuously. “The accuracy tells me when I make
changes in the index how much I
am improving the things that drive
profitability, things in the goal,”
Spangler said. “If I have more EPDs
for economically relevant traits, that
improves the accuracy of the index.
When we include more economically
relevant traits and increase the
accuracy of the EPDs, we increase
the accuracy of the index.”
“The selection indexes are fairly robust
against genetic variance, but if we
happen to use incorrect parameters,
then those would need to be updated
also,” Spangler explained. “Also, if the
inference of a trait changes, then we
WHAT AFFECTS SELECTION DECISIONS?
Changes to … • trait definitions (scaling)
• goal traits • economic parameters/
assumptions
• traits with EPD (index traits)
• genetic co-variances
• component trait accuracy
12 | AUGUST 2018
• population (assumed) means
Source: Matt Spangler, BIF 2018
Some traits may have a different
scale and inference, examples being
marbling and reproductive longevity.
The definition for those traits may
change, and so the means would
need to be re-evaluated because
their weight on the index would
possibly increase or decrease.
“Sensitivity is determined by weight in
the index,” Spangler said. “This single-
step process increases accuracy of the
selection index. We need to remember
that re-ranking is not a bad thing if it
moves us closer to the truth. Something
we still need to figure out is when
reasonableness checks are necessary.”
Some believe a universal index for
use across all breeds should be
the next step. However, Spangler